You are on page 1of 38

FACE RECOGNITION BASED

ATTENDANCE SYSTEM
Amal Chandran
S4 MCA
Guided by:
Mrs Shemeema Hashim
Overview
• INTRODUCTION
• LITERATURE SURVEY
• COMPARISON TABLE
• PROPOSED SYSTEM
• GANTT CHART
• CONCLUSION
INTRODUCTION
• Traditional attendance methods are outdated and prone to errors.
• Face recognition technology provides an efficient solution using
unique facial features.
• The main goal is to automate attendance, reducing manual efforts
and errors.
• The system uses facial recognition for real-time, high-accuracy
attendance tracking.
• It offers a user-friendly interface for easy adoption by administrators
and users.
• Components include high-resolution cameras, facial recognition
algorithms, and database integration.
LITERATURE SURVEY
1. Face Recognition based Attendance Management System(2020)
• The system consists of four phases: database creation, face detection,
face recognition, and attendance updation.
• Dataset: Images of students are captured using a web camera, pre-
processed, and saved as the names of respective students in a folder.
• Face Detection: Haar-Cascade Classifier with OpenCV is used for face
detection.
• Face Recognition: Local Binary Pattern Histogram algorithm is used for
face recognition.
• After face recognition process, the recognized faces are marked as
present in the excel sheet, and the rest are marked as absent.
Face Recognition based Attendance Management System(2020)
• Advantages:
➢Automation: The system automates the attendance marking process, reducing the time
and effort required for manual attendance taking.
➢Efficient Real-time Attendance Tracking:Face recognition technology ensures accurate
attendance while providing real-time updates, minimizing the risk of proxy attendance

• Disadvantages:
➢Accuracy and Reliability: The accuracy and reliability of face recognition technology can
be influenced by factors such as lighting conditions, image quality
➢challenges in Implementation: demands technical expertise and substantial resources.
Cost considerations, encompassing hardware, software, and training, pose potential
challenges for institutions with budget constraints.
2. Development of an Automatic Class Attendance System using CNN-
based Face Recognition (2020)
• The system uses Convolutional Neural Networks (CNN) for face
recognition
• By leveraging CNN technology, the system can detect and recognize
multiple faces from a video stream, automatically recording daily
attendance with an average recognition accuracy of about 92%.
• Continuous photos of enrolled students are captured from a live
video stream, along with their names and IDs, to create a dataset for
training.
• The dataset is trained using a CNN model to recognize facial features
and identities
Development of an Automatic Class Attendance System using CNN-
based Face Recognition (2020)
• Advantages:
➢Enhanced Efficiency: The system streamlines the attendance process, reducing
the time and effort required for manual attendance tracking.
➢Automation: The system automatically detects and recognizes multiple faces
from a video stream, simplifying the attendance recording process.
• Disadvantages:
➢Hardware and Infrastructure Requirements: The implementation of the system
may require specific hardware and infrastructure, such as high-performance
computers and well-positioned cameras.
➢Training Time: The time taken to train the dataset increases with the size of the
dataset
3. Automated Student Attendance System Using Face Recognition
(2020)
• The system presents an automated student attendance system using
face detection algorithms, deep learning for student identification.
• The system is implemented with a graphical user interface using
Python, aiming to streamline attendance management.
• Histogram of Oriented Gradients and Haar-Cascade algorithms, to
detect human faces in digital images.
• For face recognition, deep learning based on convolutional neural
networks (CNNs) is employed, along with a support vector machine
(SVM) classifier.
• The Histogram of Oriented Gradients (HOG) algorithm achieved 100%
precision and recall, while the Haar-Cascade algorithm had a
precision of 69%.
Automated Student Attendance System Using Face Recognition (2020)

• Advantages:
➢The use of a GUI enhances the usability and accessibility of the system for
instructors.
➢he paper addresses the important issue of automated student attendance using
innovative technologies like face detection algorithms and deep learning.
• Disadvantages:
➢Limited information on the training process of the deep learning model, making it
difficult to assess the robustness and reliability of the system.
➢Potential challenges in real-world implementation, such as scalability to larger
classroom sizes or variations in environmental conditions, may hinder the
practical usability of the model.
4. Comparative Study of Feature-based Algorithms and Classifiers in
Face Recognition for Automated Attendance System(2020)
• It compares three state-of-the-art appearance-based feature
extraction methods (PCA, LDA, and a Hybrid approach) under even
conditions.
• Experiments were conducted on a face database (SDB) with varying
conditions like different lighting, poses, and expressions.
• Viola Jones algorithm is employed for face detection, known for its
reliability, robustness, and real-time applicability.
• System Utilizes feature extraction methods such as PCA, LDA, and a
Hybrid approach for recognizing faces in the testing dataset.
• The accuracy of the face recognition system was evaluated using the
False Rejection Ratio (FRR)-PCA(90.1%),LDA(92.3%) & Hybrid(95.6%).
Comparative Study of Feature-based Algorithms and Classifiers in Face
Recognition for Automated Attendance System(2020)
• Advantages:
➢Hybrid Approach: The hybrid approach, combining PCA, LDA, and SVM, is
highlighted as an improvement over individual methods, potentially leading to a
higher face recognition rate.
➢Real-Time Application: The system is designed for real-time applications
• Disadvantages:
➢The system's performance for recognizing a large number of faces in an input
image (more than 50 or 100) is mentioned as an area for improvement.
➢The system's recognition rate may decrease when faced with challenges such as
dark or cloudy lighting conditions, indicating a potential limitation in certain
environmental settings.
5. Face Recognition Based Automated Attendance Management
System(2022)
• Design methodology: used a descriptive framework to design the
proposed system, which utilizes RFID technology and the SURF
algorithm to create a student attendance control system.
• Algorithm selection: uses the OpenCV algorithm for face recognition
and the Viola-Jones technique for face detection.
• Hardware selection: used a Raspberry Pi and camera module to
capture and store images in a database.
• Performance evaluation: evaluated the performance of the system by
measuring the accuracy and speed of face recognition and
attendance tracking.
Face Recognition Based Automated Attendance Management
System(2022)
• Advantages:
➢The system can prevent proxy attendance and improve security.
➢The system can be integrated with other technologies, such as RFID and NFC.
➢Automated attendance tracking saves time and reduces the workload of faculty.
• Disadvantages:
➢The system requires a high level of technical expertise to develop and maintain.
➢The system may not be effective in low-light or crowded environments.
6. IoT based Intelligent Attendance Monitoring with Face Recognition
Scheme(2021)
• By utilizing face recognition technology and IoT infrastructure, the
system seeks to automate the attendance process, reduce manual
intervention, and enhance the accuracy and reliability of attendance
records.
• The raspberry-pi module is used for face detection and recognition,
and the images are compared with stored images to mark
attendance.
• The system utilizes IoT infrastructure to enable real-time data transfer
and communication between the raspberry-pi module and the cloud
platform.
• The attendance data is stored in a database management system, The
system uses MYSQL for database management.
IoT based Intelligent Attendance Monitoring with Face Recognition
Scheme(2021)
• Advantages:
➢User-friendly Interface: The web application is designed to be user-friendly and
accessible from any device with an internet connection.
➢Accurate Attendance Tracking: The system uses face recognition technology,
which ensures accurate attendance tracking for students and faculty members.

• Disadvantages:
➢Cost: The system requires the purchase of hardware and software, which can be
expensive.
➢Technical Expertise: The system requires technical expertise to set up and
maintain, which may be a challenge for some educational institutions.
7. Student attendance with face recognition (LBPH or CNN) (2023)
• Includes the use of LBPH and CNN algorithms in improving student
attendance tracking systems.
• The dataset includes images of students' faces for training and testing
the face recognition algorithms.
• The dataset likely consisted of a collection of images representing
different individuals, with variations in lighting conditions, facial
expressions, and poses to enhance the robustness of the algorithms.
• The CNN algorithm was found to have superior accuracy compared to
LBPH, with CNN achieving a higher accuracy rate of 98% 4.
Additionally, the LBPH algorithm was reported to have an accuracy
rate of 77.55%
Student attendance with face recognition (LBPH or CNN) (2023)

• Advantages:
➢Face recognition technology can help prevent fraudulent attendance practices,
such as proxy attendance.
➢Cost-Effectiveness: The required hardware, such as a webcam, is relatively
inexpensive, making the implementation of face recognition technology cost-
effective for educational institutions

• Disadvantages:
➢Dataset Requirements: Implementing CNN algorithms requires a large number of
datasets for training, which can be resource-intensive and time-consuming .
➢External Factors: The accuracy of face recognition systems can be influenced by
external factors such as lighting, face position, and background, which may
impact the system's performance
8. Research on Multi-Angle Face Detection Method Based on Improved
YOLOV2 Algorithm (2021)
• The YOLOV2 algorithm is chosen for its higher accuracy and faster
detection speed in target real-time detection.
• By integrating ResNet and optimizing the loss function, the proposed
method aims to improve accuracy and robustness in face detection.
• ResNet is introduced to improve small face target detection within
the basic structure of YOLOV2.
• The Wider Face dataset is used, containing a variety of face images
with different sizes, poses, and expressions etc.
• YOLOV2 algorithm achieved an accuracy of 89.8% on the Wider Face
dataset.
Research on Multi-Angle Face Detection Method Based on Improved
YOLOV2 Algorithm (2021)
• Advantages:
➢Improved Accuracy: The system demonstrates enhanced accuracy, as indicated by
the 5.1% improvement in face detection on the Wider Face dataset.
➢Integration of ResNet: The incorporation of the residual network (ResNet)
contributes to improved small face target detection, addressing limitations
present in traditional face detection models.
• Disadvantages:
➢The integration of ResNet and modifications to the loss function may introduce
increased complexity to the algorithm, potentially affecting computational
efficiency.
➢The system's performance could be sensitive to hyperparameter settings, and
finding optimal configurations may require thorough experimentation.
9. Facial Recognition Attendance Monitoring System using
Deep Learning Techniques (2023)

• This paper introduces a Facial Recognition Attendance Monitoring


System using Deep Learning Techniques.
• By employing advanced algorithms like Haarcascade and Eigen values,
the system detects and verifies faces against a student database.
• The system utilizes Haar Cascade classifiers for face detection and the
LBPH algorithm for feature extraction.
• OpenCV, an open-source computer vision library, is utilized for
efficient image processing and algorithm implementation.
Facial Recognition Attendance Monitoring System using
Deep Learning Techniques (2023)
• Advantages:
➢ Integration of sophisticated algorithms enhances accuracy in face detection and
recognition.
➢Real-time Processing: Utilization of OpenCV ensures efficient real-time image
processing.
• Disadvantages:
➢implementing facial recognition systems might incur high costs, especially for
organizations with limited budgets.
➢Face recognition systems may encounter difficulties in accurately identifying
individuals under varying conditions, such as changes in lighting, facial
expressions, or when individuals wear accessories.
10. ATTENDANCE SYSTEM USING CNN FACE RECOGNITION
ALGORITHM(2023)

• The system introduces an Attendance System utilizing a Face


Recognition Algorithm based on CNN.
• The process involves capturing facial images using a quality camera,
using histogram of oriented gradient (HOG) for detection, and
implementing deep learning for recognition.
• the dataset used is trained by capturing images under diverse
conditions to improve accuracy.
• The use of computer vision algorithms extracts facial features for
identification
ATTENDANCE SYSTEM USING CNN FACE RECOGNITION
ALGORITHM(2023)
• Advantages:
➢ Efficiency and Accuracy: The system offers a more efficient and accurate
attendance tracking methods, reducing the risk of errors or fraud.
➢By automating the attendance recording process, the system saves time and
effort, eliminating the need for manual data entry or verification.
• Disadvantages:
➢ Different face recognition algorithms have intricacies, and the choice of
algorithm can impact accuracy, particularly under varying conditions like lighting
or facial expressions.
➢Accuracy is dependent on the quality and diversity of training data, and
insufficient representation may impact recognition.
Comparison Table
PAPER METHOD ADVANTAGE DISADVANTAGE

Face Recognition based Uses Haar Cascade algo for Efficient Real-time Challenges in
Attendance Management Face detection and Local Attendance Tracking: Implementation: the
System(2020) Binary Pattern histogram accurate and real-time system poses challenges
for face recognition attendance tracking. that demand technical
expertise and substantial
resources. Also high cost
Development of an The dataset is trained Enhanced Efficiency: Training Time: One
Automatic Class using a CNN model to improves the efficiency by drawback is the increased
Attendance System using recognize facial features automating the detection time needed to train the
CNN based Face and identities and recognition of multiple dataset, especially with a
Recognition faces from a video stream. larger dataset
(2020)
Automated Student Histogram of Oriented The use of a GUI enhances Limited Information on
Attendance System Using Gradients and Haar- the system's usability and Training Process: The lack
Face Recognition (2020) Cascade algorithms, to accessibility for instructors. of detailed information on
detect human faces in the training process of the
digital images. deep learning model poses
a disadvantage.
PAPER METHOD ADVANTAGE DISADVANTAGE

Comparative Study of Viola Jones algorithm is The hybrid approach, The system's performance
Feature-based Algorithms employed for face combining PCA ,LDA, and might face challenges when
and Classifiers in Face detection and uses feature SVM has the potential to recognizing a large number
Recognition for Automated extraction methods such as achieve a higher face of faces in an input image,
Attendance System (2020) PCA, LDA, and a Hybrid recognition rate, indicating especially exceeding 50 or
approach for recognizing improved accuracy and 100 faces.
faces in the testing dataset. effectiveness.
Face Recognition Based uses the OpenCV algorithm The system has the The development and
Automated Attendance for face recognition and the capability to prevent proxy maintenance of the system
Management System(2022) Viola-Jones technique for attendance, improving demand a high level of
face detection. overall security within the technical expertise.
attendance management
process.
IoT based Intelligent raspberry-pi module is used IoT infrastructure enhance This cost factor can be a
Attendance Monitoring for face detection and the accuracy and reliability limitation for purchase of
with Face Recognition recognition. of attendance records. hardware and software.
Scheme(2021)
PAPER METHOD ADVANTAGE DISADVANTAGE

Student attendance with use of LBPH and CNN Cost-Effectiveness: The Implementing CNN
face recognition (LBPH or algorithms in improving required hardware, like a algorithms, while effective,
CNN) student attendance webcam, is relatively demands a substantial
(2023) tracking systems. inexpensive. number of datasets for
training.

Research on Multi-Angle The YOLOV2 algorithm is The system showcases The integration of ResNet
Face Detection Method used and ResNet is enhanced accuracy in face and modifications to the
Based on Improved YOLOV2 introduced to improve detection on the Wider loss function may introduce
Algorithm (2021) small face target Face dataset increased complexity to the
algorithm.

Facial Recognition Haar Cascade classifiers Increased accuracy as Implementing facial


Attendance Monitoring used for face detection and integration of algorithms recognition systems may
System using Deep the LBPH algorithm for incur high costs
Learning Techniques (2023) feature extraction.
PAPER METHOD ADVANTAGE DISADVANTAGE

ATTENDANCE SYSTEM using histogram of oriented The system provides a the choice of algorithm can
USING CNN FACE gradient (HOG) for more efficient and accurate impact accuracy,
RECOGNITION detection, and attendance tracking particularly under varying
ALGORITHM implementing deep method conditions such as lighting
(2023) learning for recognition. or facial expressions
Proposed System
Proposed System
• The proposed system aims to integrate the strengths of existing face
recognition-based attendance management systems, addressing key
considerations.
• the system will utilize an advanced combination, possibly integrating
Haar-Cascade for robust face detection and LBPH for accurate face
recognition. The algorithm choice will prioritize reliability under
varying conditions.
• the system will implement a diverse training dataset captured under
different lighting conditions, facial expressions, and poses. CNN will
be employed to enhance feature extraction and recognition accuracy.
Gantt Chart
Conclusion
Conclusion

• Face recognition technology enhances efficiency, accuracy, and


security in attendance management.
• A proposed system integrates strengths of existing approaches,
prioritizing reliability, efficiency, and scalability.
• The system uses robust face detection algorithms and advanced
recognition techniques.
• The system aims to meet evolving needs of educational institutions
through careful algorithm selection, diverse dataset collection, and
ongoing refinement.
Reference
1. Smitha, Pavithra S Hegde, Afshin “Face Recognition based Attendance
Management System(2020)” IJERT 2020
https://doi.org/10.22214/ijraset.2022.41237
2. Soumitra Chowdhury, Sudipta Nath, Ashim Dey and Annesha Das “Development
of an Automatic Class Attendance System using CNN-based Face Recognition ”
IEEE 2020 https://doi.org/10.1109/ETCCE51779.2020.9350904
3. E. Omer Akay, K. Oguz Canbek, Yesim Oniz “Automated Student Attendance
System Using Face Recognition” IEEE 2020
https://doi.org/10.1109/ISMSIT50672.2020.9255052
4. Ms.Sarika Ashok Sovitkar, Dr. Seema S. Kawathekar “Comparative Study of
Feature-based Algorithms and Classifiers in Face Recognition for Automated
Attendance System” IEEE 2020
https://doi.org/10.1109/ICIMIA48430.2020.9074917
5. Aparna Trivedi, Chandan Mani Tripathi, Dr. Yusuf Perwej, Ashish Kumar
Srivastava, Neha Kulshrestha “Face Recognition Based Automated Attendance
Management System ” IJSRST 2022
https://doi.org/10.32628/IJSRST229147
6. S A Sivakumar, G Thamarai Selvi, C Udhaya Shanka, Tegil J John, Bhukya Madhu,
Arjun K P “IoT based Intelligent Attendance Monitoring with Face Recognition
Scheme ” IEEE 2021 https://doi.org/10.1109/ICCMC51019.2021.9418264
7. Andre Budimana, Fabiana, Ricky Aryatama Yaputeraa, Said Achmada, Aditya
Kurniawan “Student attendance with face recognition (LBPH or CNN)” 2023
http://dx.doi.org/10.1016/j.procs.2022.12.108
8. Huanping Zhang, Hanhua Cao, Zhendan Liu, Yuhuai Zhou and Yujuan Wang
“Research on Multi-Angle Face Detection Method Based on Improved YOLOV2
Algorithm ” IOP 2021 http://doi.org/10.1088/1742-6596/1848/1/012024

9. Dr A Manjula, D. Kalpana, Sanjay Guguloth “Facial Recognition Attendance


Monitoring System using Deep Learning Techniques” IJIEMR 2023
http://www.ijiemr.org/downloads.php?vol=Volume-12&issue=Issue
10. Sudarshan R. Date, Aditya S. Chikane, Soham M. Dabake, Mohan K. Mali
“ATTENDANCE SYSTEM USING CNN FACE RECOGNITION ALGORITHM ” IJCRT 2023
http://www.ijcrt.org/papers/IJCRT2304755.pdf
THANK YOU

You might also like